The WASD package and updates will always be distributed as ZIP archives.
For complete package distributions the archive will contain the complete directory tree. Hence to install or do a complete update it is necessary to SET DEFAULT into the top-level directory of the disk the package is to be installed on.
$ SET DEFAULT device:[000000]
For updates to portions of the package only the tree below HT_ROOT:[000000] is provided, hence it is necessary to SET DEFAULT to HT_ROOT:[000000] before UNZIPping.
$ SET DEFAULT HT_ROOT:[000000]
In either case it is recommended to check the integrity of, then list the contents of, the archive before UNZIPping. This may be done as follows:
$ UNZIP -t device:[dir]archive.ZIP $ UNZIP -l device:[dir]archive.ZIP
A complete installation/update (the former) will have the structure:
Archive: HT_ROOT:[000000]HTROOT430.ZIP;1 WASD (HFRD) VMS Hypertext Services, Copyright (c) 1996,1997 Mark G.Daniel. This package (all associated programs), comes with ABSOLUTELY NO WARRANTY. This is free software, and you are welcome to redistribute it under the conditions of the GNU GENERAL PUBLIC LICENSE, version 2. * Complete v4.3.0 package. * $ SET DEFAULT device:[000000] * $ UNZIP "-V" device:[dir]HTROOT430.ZIP VMS file attributes saved ... use UnZip 5.n+ on OpenVMS Archive created 1-AUG-1997 Length Date Time Name ------ ---- ---- ---- 0 07-30-97 14:15 ht_root/$_read_1st/ 0 07-30-97 14:15 ht_root/aacrt060/ 0 07-30-97 14:15 ht_root/auth/ 0 07-30-97 14:15 ht_root/axp/ 5360 07-30-97 13:23 ht_root/changes.html 0 07-30-97 14:15 ht_root/doc/ . . . 16896 07-30-97 09:23 ht_root/vax/wwwrkout.exe 16384 07-30-97 09:21 ht_root/vax/wwwrkout_netlib.exe ------ ------- 19109754 996 files
An update (the latter) will have the structure:
Archive: HT_ROOT:[000000]HTROOT430_UPD_01.ZIP;1 WASD (HFRD) VMS Hypertext Services, Copyright (c) 1996,1997 Mark G.Daniel. This package (all associated programs), comes with ABSOLUTELY NO WARRANTY. This is free software, and you are welcome to redistribute it under the conditions of the GNU GENERAL PUBLIC LICENSE, version 2. * Mostly minor updates to scripts and utilities for the v4.3.0 package. * A couple nuisance-value-bugs have also been fixed in these. * SDM2HTM can now generate <FRAMESET> documents (see [DOC]*.COM) VMS file attributes saved ... use UnZip 5.n+ on OpenVMS Archive created 6-AUG-1997 Length Date Time Name ------ ---- ---- ---- 42957 08-04-97 07:57 src/utils/wwwrkout.c 59664 08-04-97 07:57 src/utils/obj_axp/wwwrkout.obj 19674 08-04-97 07:57 src/utils/obj_vax/wwwrkout.obj 26624 08-04-97 07:57 axp/wwwrkout.exe 27136 07-30-97 09:21 axp/wwwrkout_netlib.exe . . . 45056 08-05-97 11:02 vax/sdm2htm.exe 617 08-05-97 09:23 doc/all_docs_framed.com 44 08-05-97 09:23 doc/all_docs_not_framed.com 544 08-05-97 09:22 doc/readme.html ------ ------- 2773107 90 files
The archive may then be UNZIPped. If this is a completely new installation a new directory tree will be created. If an update to an existing installation the tree will be created "over-the-top" of the existing structure.
When updating:
$ BACKUP HT_ROOT:[000000...] location:HTROOT.BCK/SAVE/VERIFY $ ZIP "-V" location:HTROOT.ZIP device:[HT_ROOT...]*.* $ ZIP "-T" location:HTROOT.ZIP
The package comes complete with both VAX and Alpha executables, object modules and complete sources. To help conserve disk space, the HT_ROOT:[SRC]PRUNE.COM procedure is provided.
If it is a stand-alone system, or a cluster containing only the one hardware architecture, either VAX or Alpha, the executables for the other can be deleted using:
$ @HT_ROOT:[SRC]PRUNE EXE
If it is suspected that a re-link will not be necessary (which is unlikely unless extensive re-development is to be undertaken), all object files may be deleted using the following DCL:
$ @HT_ROOT:[SRC]PRUNE OBJ
Being even more ruthless, the entire source tree (C, build and object files) can be eliminated using:
$ @HT_ROOT:[SRC]PRUNE SRC
Any more and you've decided to use some other server package, so get rid
of it yourself! :^)
4.1 - Local Setup Suggestions
Package updates will never contain anything in these three directories:
To prevent the overwriting of local configuration files it is suggested these be placed in the HT_ROOT:[LOCAL] directory. Startup files will most probably be placed where-ever the local site manages system startup. These could be placed in the HT_ROOT:[LOCAL] directory if there is nowhere more appropriate. Authentication databases could also be placed in the [LOCAL] directory. The server account's support files (LOGIN.COM, etc.) may be copied from the HT_ROOT:[EXAMPLE] directory into HT_ROOT:[HTTP$SERVER] and customized as necessary.
In like manner, local DCL scripts or DCL wrappers for executable scripts should be placed in HT_ROOT:[SCRIPT_LOCAL]. Executables should continue to be placed in HT_EXE:
Changes that are made in package configuration files, etc., can then be
propagated to the local areas as appropriate.
4.2 - Organizing Documents/Data
It is recommended that the server distribution tree and any document and other Web-specific data areas be kept separate and distinct. The former in HT_ROOT:[000000], the latter in something like WEB:[000000]. This logical device could be provided with the following DCL introduced into the server startup procedures:
$ DEFINE /SYSTEM /TRANSLATION=CONCEALED WEB DSA811:[WEB.]
The logical organisation of served data is largely hierarchical, organised under the Web-server path root, and is achieved via two mechanisms.
WASD has a single section of the file system for the core Web files, such as the Division home page, help areas, Web documentation, etc., it can be accessed using the logical device WEB:[000000]. Incorporated with this are other subdirectories providing storage for specific collections of data, such as committee minutes, administration announcements, etc.
Physically distinct areas are used for good physical reasons (e.g. the area can best be hosted on a task-local disk), for historical reasons (e.g. the area existed before any Web environment existed) or for reasons of convenience (e.g. lets put this where access controls already allow the maintainers to manage it).
Reasons for an area being physically integrated with the core Web data
area can be legitimate (e.g. there is really nowhere else it reasonably
belongs), convenience (e.g. let's quickly put it here) or by logical necessity
(it really does below as part of the core Web enviroment, e.g. documentation
on HTML, etc.)
Guidelines
In general, only Web-related files need to go into the core physical Web file area. All other groupings should, if possible, be decentralised into the portion of the file system they represent and logically placed in the server's path using rules in the mapping file. That is, a given project's Web files should be located in the project's part of the file system. If it doesn't have any then it may be a candidate for location in the core area.
When locating a Web area in a physically distinct area it is possible the maintainers of that data will already have the correct access controls. If locating an area in the core hypertext environment it will be necessary to give the mantainer ownership of the directory area (and possibly disk quota), or provide ACL access if multiple maintains are involved.
When locating Web-accessable data in a physically distinct area it will be
necessary to update the mapping file with a new rule (see
8 - Mapping Rules). If located within the core Web area the
rules do not need to be adjusted.
4.3 - Reporting Problems
This package, as is generally the case with freeware, is mainly developed and supported outside of the author's main occupation and working hours. Reports of problems and bugs (while not necessarily welcome :^), as well as general queries, are responded to as soon a practicable. If the documentation is inaccurate or could benefit from clarification in some area please advise of this also (the better the documentation the less queries you have to field personally ... or so the theory goes).
With all reports please include the version of the server or script, and the hardware platform, operating system and TCP/IP package and version in use.
If a server error message is being generated please examine the HTML source of the error page. The "<META...>" information contains version information as well as valuable source code module and line information. Include this with the report.
If the server is exiting with an server-generated error message this information also contains module and line information. Please include this with the report.
Image crash dumps may also be generated, although these are of less value than the case of the previous two.
Reports may be e-mailed to Mark.Daniel@dsto.defence.gov.au.